Oil-water separation and flotation

Why fine oil droplets escape conventional flotation

Conventional flotation loses the smallest oil droplets because they follow the water around the bubbles instead of meeting them. Dissolved air flotation (DAF) releases air from pressurized water as microbubbles, typically 30 to 100 µm across. DAF separates free oil and larger suspended solids well but often struggles with emulsified droplets smaller than 20 µm, which travel along the streamlines around a rising microbubble, so collision efficiency is low Shen et al., 2022.

A nanobubble is smaller than 1 µm; those used in oily-water flotation are typically 150–350 nm. At that size buoyancy is weak and Brownian motion dominates, so nanobubbles stay suspended in the water column for long periods. The review by Shen et al., 2022 links this to a much higher probability of collision with the ultrafine droplets that DAF misses.

How nanobubbles attach to oil droplets

Nanobubbles attach to oil because oil is hydrophobic, and once attached they make larger bubbles stick more readily.

• Nucleation on hydrophobic surfaces. Nanobubbles form preferentially on hydrophobic surfaces such as oil droplets, where the energy barrier is lower. Once attached, they form a gas bridge between droplets or act as a secondary collector.
• Larger contact angle. Attached nanobubbles increase the apparent contact angle of droplets and particles and extend the three-phase (gas, liquid, solid) contact line. The aggregate becomes more hydrophobic, and larger carrier microbubbles then attach to it and lift it more stably to the surface.

This "seeding" let nanobubbles turn hydrophilic or weakly hydrophobic impurities into aggregates that could be separated Oliveira et al., 2017; Shen et al., 2022.

Emulsion breaking with lower reagent doses

Nanobubbles worked alongside demulsifiers and polymers in the studies, lowering the doses needed rather than replacing them. Stable oil-in-water emulsions usually need high doses of chemical demulsifiers to overcome the electrostatic repulsion between droplets, which is measured as zeta potential. Three physical effects were reported.

• Aerated flocs. Nanobubbles became trapped inside flocculated oil structures instead of only attaching to the outside. They lowered floc density without heavy polymer bridging and produced buoyant "aerated flocs" that resisted shear as they rose Etchepare et al., 2017.
• Lower surface tension. Nanobubbles lowered the surface tension of the water by about 7–20%, depending on the gas. This helped oil droplets coalesce and surfactants or polymers adsorb, so lower chemical doses achieved phase separation English, 2025.
• Zeta potential. Bubbles generally carry a negative charge. Their large specific surface and their interaction with ions in saline water can modulate the zeta potential of the system, reducing repulsion between droplets and helping them aggregate Shen et al., 2022.

Reported removal results

Adding nanobubbles to flotation gave higher oil removal than the conventional comparison in each study below, including in saline water.

• Emulsified crude oil in saline water. With initial oil concentrations of 334–484 mg/L, micro- and nanobubbles together removed more than 99% of the oil, leaving less than 1 mg/L. Even at a lower saturation pressure of 3.5 bar, the effluent met the offshore discharge standard of less than 29 mg/L Etchepare et al., 2017.
• Petroleum hydrocarbons in sand. In oil-contaminated sand and fluids, positively charged nanobubbles removed at least 94% of total petroleum hydrocarbons (TPH) under intermittent injection, more than under batch injection Bui et al., 2022.
• Flotation against a sparging control. Oil removal rose from 62% with standard air sparging to 98.5% with air-nanobubble flotation at an optimized polymer dose of 0.2 wt%, and the nanobubble process reached its result roughly 20% faster English, 2025.
• Oil and grease in municipal effluent. Oxygen nanobubbles reduced oil and grease in secondary effluent by 33%, from 9 to 6 mg/L, while air bubbles showed no significant effect over the same period Ahmed et al., 2023.